RF linear amplifiers utilize devices that exhibit non-linear characteristics at higher power levels whereby signal distortion is introduced. For example, if more than one carrier signal is applied to a linear amplifier, its non-linear characteristics cause an unwanted multiplicative interaction of the carrier signals being amplified and the amplifier output contains intermodulation products or distortion. These intermodulation products cause interference which may exceed established transmission standards.
As is well known, intermodulation distortion can be reduced by separating the distortion component of the amplifier output and feeding forward the distortion component to cancel the distortion in the amplifier output signal. FIG. 1 is a simplified block diagram of a feed forward circuit 10 disclosed in U.S. Pat. No. 4,885,551. Feed forward circuit 10 receives input signal S having at least one carrier in a prescribed frequency range. Input signal S is split into signals S(12a) and S(12b), wherein signals S(12a) and S(12b) are signals representative of the input signal S. For ease of discussion, numeric references in parenthesis are used herein to indicate from which component(s) a signal was output, and alphanumeric references in parenthesis are used herein to indicate that the component(s) from which the signal was output has more than one output. For example, signal S(12a) would indicate that it was an output signal of splitter 12 and that it was one of a multitude of output signals from splitter 12. If a signal has more than one reference, the order of the references would indicate the path of the signal. For example, signal S(12a,14) would indicate that it was first an output signal of the splitter 12 and then an output signal of an amplifier 14. The output signal of the latter referenced component would be a signal representative of the preceding referenced component, e.g., signal S(12a,14) is a signal representative of the signal S(12a).
Signal S(12a) is applied to a first or main circuit path having main amplifier 14 which amplifies signal S(12a) and introduces distortion signal D(14). Thus, main amplifier 14 produces output signal S(14) comprising signals S(12a,14) and D(14). Signal S(14) is applied to directional coupler 18 which directs signals S(18a) and S(18b) to delay 22 and cancellation circuit 20, respectively, wherein signal S(18a) comprises signals S(12a,14,18a) and D(14,18a) and signal S(18b) comprises signals S(12a,14,18b) and D(14,18b). Signal S18(a) is delayed by delay 22 to produce output signal S(22) comprising S(12a,14,18a,22) and D(14,18a,22).
Signal S(12b) is applied to a second circuit path where it is delayed by delay 16 to produce output signal S(16) comprising signal S(12b,16). Signal S(12b,16) is combined with signal S(18b) in cancellation circuit 20 to form output signal S(20). In cancellation circuit 20, signal S(12a,14,18b) (via signal S(18b)) is canceled by signal S(12b,16). Thus, signal S(20) comprises distortion signal D(14,18b,20). The signal S(20) is applied to correction amplifier 24 which amplifies signal S(20) and introduces distortion signal D(24). Note that distortion signal D(24) is approximately 10,000 times smaller in amplitude than distortion signal D(14). Thus, correction amplifier 24 produces output signal S(24) comprising distortion signals D(14,18b,20,24) and D(24). Signal S(24) is combined with signal S(22) in cancellation circuit 26 to produce output signal S(26). The amplitude and phase of distortion signal D(14,18b,20,24) (via signal S(24)) should be approximately equal to the amplitude and phase of distortion signal D(14,18a,22) (via signal S(22)) such that distortion signals D(14,18b,20,24) and D(14,18a,22) cancel each other in cancellation circuit 26. Thus, signal S(26) comprises S(12a, 14,18a,22,26) and D(24,26). In effect, the amplitude of the distortion signal in output signal S(26) is being reduced by substituting a signal representative of distortion signal D(14) (i.e., D(14,18a,22)) with distortion signal D(24), which has a smaller amplitude.
Ideally, correction amplifier 24 should apply a constant amplitude gain and/or phase shift to signal S(20) such that the amplitude and phase of distortion signal D(14,18b,20,24) (via signal S(24)) should always be approximately equal to the amplitude and phase of distortion signal D(14,18a,22) for purposes of achieving good cancellation. However, as conditions change over time, such as temperature, the amplitude gain and/or phase shift applied to signal S(20) by correction amplifier 24 may vary. Such variation may adversely affect the amount distortion signal D(14,18b,20,24) cancels distortion signal D(14,18a,22). Therefore, there exists a need to compensate for changes in conditions to the correction amplifier such that good cancellation of distortion signals is maintained.